You are on page 1of 8



Deborah Weaver

Institute of Linguistics

Michif is a dialect or language growing out of the contact betw

Europeans and Native Americans, primarily French and Cree. It is an
unusual if not unique linguistic product, with its entire noun phrase,
except for a few Cree nouns, coming from French, and its verb phrase
and overall syntax coming from Cree, albeit with considerable French
influence. Spoken by residents and dependents of the Turtle Mountain
Reservation (who are primarily Metis) in north-central North Dakota,
as well as in parts of Canada, it is a language undergoing not only coalescence and change, but also death due to increasing English monolingualism of Michif children.
Research in Turtle Mountain Michif has been done primarily by
John Crawford (1976, 1978) and graduate students working with him.
Michif wasfirstintroduced to the Algonquian conference in 1976 by
Richard Rhodes in his paper "French Cree: a Case of Borrowing"
(Rhodes 1977). The purpose of the present paper is to explore the
effect of language change and death on one characteristic feature of
Algonquian languages, obviation, as exhibited in the speech of the residents of Turtle Mountain. Obviation is difficult to discover in Michif
because, while it has preserved most of the distinctions of Plains Cree
in its verb paradigms, its noun phrase is French. It could be expected
that with the loss of most of the Cree nouns the cross-referencing system, including obviative inflection, would fail to transfer over to the
French nouns, making it difficult to elicit obviative subject forms in
the verb paradigm. With a lack of obviative inflection on nouns, one
might suppose that the corresponding forms could have been lost in
verbs. However, obviative inflection does occasionally occur, especially
on the few existing Cree nouns, making the question of what factors
are affecting that occurrence one worthy of exploration.
More significant than the inherent structure of the language in making obviative forms hard to discover are the sociolinguistic factors surrounding the current use of Michif on the reservation. English has
become the main vehicle of communication and, because of this situation, the linguist often finds it difficult to get an accurate picture of


Deborah Weaver

Michif structure. Since Michif is used mainly within individual family

groups, at times it seems like each family speaks a different dialect.
English is the acceptable m o d e of communication outside the family,
making it difficult to obtain data. Thus the linguist m a y have to rely on
translations of isolated, unrelated sentences. Since English has nothing
like the proximr.te/obviative distinction, it is difficult to elicit it in such
translations. T h e novice must proceed with caution: there is no guarantee that because she has been unable to elicit a form that it does not
occur. This is true in any language learning situation, but even more
so in a situation as complex as this one.
Whether a person uses obviative forms m a y be affected by m a n y
factors. T h e degree to which speakers are able to use the language to
express themselves and interact with other people could be expected to
influence the degree to which they use features like obviation that often
serve a discourse level function. T h e same could be said for the degree
to which speakers actually use the language, which in this particular
sociolinguistic situation is often less than they are capable of. S o m e
speakers do not think of Michif as a language, but as bits and pieces
of other languages, preferring not to use it on a regular basis.
T h e degree of fluency and current use of the language m a y be expected to correlate with the degree to which morphological leveling
of several types has occurred. S o m e speakers preserve the full range
of person affixes and are most likely also to preserve obviative forms.
Others exhibit a leveling of person prefixes and tense markers and are
more likely to neutralize the proximate/obviative distinction. S o m e
speakers use Cree possessive markers on the few remaining Cree nouns
including the expected obviative endings as in (1) below:

li garso o-musum-a ki:-pakamahw-e:w

the boy his-grandfather-obv past-hit-3
'the boy hit his grandfather'

Others use the French possessive markers as in (2) below:


li garso su m u s u m ki:-pakamahw-e:w
the boy his grandfather past-hit-3
'the boy hit his grandfather'

In this study of obviation in Michif a questionnaire w a s used which

took these factors into consideration. This questionnaire sought to
establish which sociolinguistic variables might affect whether or not a
particular speaker uses obviative forms. It was anticipated that family,
age, geographical location, languages spoken in the h o m e w h e n growing

Obviation in Michif


up and presently, the types of situations in which Michif is currently

used, and perception of Michif as a language in its own right or as only
bits and pieces of other languages would be factors that could possibly
affect the use of obviative forms. This questionnaire was not intended
to be a statistical tool, but a means of identifying some of the trends
in language use on the reservation. Because of the small size of the
population that actually used Michif on a regular basis, it was deemed
impractical to seek out a statistically sound sample.
The questionnaire also served to elicit language data. The design of
this elicitation tool is crucial. Obviative forms do not occur in isolation
but as a result of two third persons being closely related syntactically.
Ideally, linguistic data should take the form of natural texts, elicited
over a period of time, from a cross-section of the population. However,
such long-term study is not always feasible, making it necessary to
design very carefully the elicitation technique used.
The part of the questionnaire reproduced in Appendix I asks for
the Michif translation of twelve English sentences. These were asked
in pairs, thefirstsentence intended to establish which third person
is proximate and which is obviative, the second reversing their initial
relationship (semantic roles).
Thefirstpair was thought to be the pair most likely to produce
an obviative marker on the noun. There are three reasons for this:
1) 'grandmother' is possessed by a third person, an environment that
makes obviation obligatory in other Algonquian languages; 2) 'grandmother' is the goal of a verb with a third person actor, an environment
that makes obviation obligatory in most Algonquian languages; and 3)
'grandmother' is one of the few remaining Cree nouns in Michif and
thus more likely to exhibit Cree morphology. It was anticipated that
thefirstsentence would establish 'grandmother' as obviative and that
when it became the actor in the second sentence it would remain obivative, causing the verb to be marked for an obviative subject. In order
to double-check the results of this, another pair of sentences with an
identical syntactic environment (2a,b) was used.
The next four sentences repeat this except that the possessed noun is
French. It was thought that these forms would be less likely to demonstrate obviation as French nouns usually do not exhibit Cree noun
morphology. The last four sentences eliminate one futher motivation,


Deborah Weaver

possession. This would be even less likely to demonstrate obviation in

an Angonquian language.
The data collected in this study are presented in Weaver (1982) along
with a complete discussion of the questionnaire and its results. I would,
however, like to discuss the way various speakers translated these sentences and how some sociolinguistic factors m a y have affected them.
The twelve speakers interviewed can be placed in three groups. The
manner in which each group translated the questions was fairly consistent, with only minor stylistic differences within each group. Examples
representative of each of these groups are given in Appendix II.
Group A included 8 of the 12 speakers interviewed. They ranged
in age from 42 to 66 and were unrelated except for a half-sister and
brother. They were born on various parts of the reservation and were
influenced by a variety of languages spoken in the home. All of them,
except one, marked third person Cree nouns in the object position as
being obviative and used the inverse form when that obviative noun
became subject. For example:


li garso ii:-miisum-a ki:-pakamahw-e:w

the boy his-grandfather-obv past-hit-3
'the boy hit his grandfather'
u:-musum-a ki:-pakamahw-ik-e:w
his-grandfather-obv past-hit-inverse-3
'his grandfather hit him'

More significantly, while only two of them ever marked a French

noun as obviative, and then only once, all of them used the inverse
form when the French noun became the subject in the second sentence
of each pair.


lafisa sor ki:-wa:pam-e:w

the girl her sister past-see-3
'the girl saw her sister'
sa sor ki:-wa:pam-ik-e:w
her sister past-see-inverse-3
'her sister saw her'

The inverse marker is required in Algonquian when a lower ranking

person in the person hierarchy of 2nd > 1st > 3rd prox. > 3rd obv.
serves as subject while a higher ranking person serves as object. The
use of the inverse marker in sentences with French nouns would seem to
indicate that the concept of another third person is, at least for these
speakers, not dependent on a noun being morphologically obviative. It
is obviative by virtue of its relationship to the other participant in the

Obviation in Michif


first sentence and thus requires the inverse marker when functioning as

Group B consists of only two speakers, ages 63 and 26, from different parts of the reservation. They retain Cree possessive markers
and obviative markers on the Cree nouns but do not spontaneously use
the inverse marker in the second sentence of each pair. The 63 year
old did, however, use inverse forms when prompted by his wife. He
finally ceased translating the sentences at all and let his wife, whose
answers mostly corresponded with those in Group A,finish.For these
two speakers, then, there are vestiges of obviation left in their speech,
but no concept of proximate and obviative forms being at different levels in the person hierarchy. Possibly they view the obviative ending as
being the second half of a split morpheme marking third person possession.
Finally, in Group C, the three speakers, ages 26, 40, and 46, do
not mark obviation without prompting.1 In their speech, Cree nouns
have been completely incorporated into the French noun phrase and
they consistently use direct forms in all sentences involving two third
The two factors, besides age, which seem to be most crucial in affecting retention of the proximate/obviative distinction are the degree
to which the language is currently used by the subjects and their attitude towards it. These are probably aspects of the same phenomenon.
Speakers w h o perceive Michif as being less than a language, or who
see themselves as unable to speak any language well, did not exhibit a
semantic concept of obviation.
A good example of the effect of attitude on speech is the 63 year old
from Group B. He grew up speaking the language, his mother spoke
only Michif, his wife had to become morefluentin order to live with his
people, and yet he almost completely blanks out when asked to speak.
W h y is this? It is impossible to say what all the factors are that led to
his saying, "I speak a little French, a little Cree, a little Chippewa, a
little English, but I don't speak any language well." W h e n I said, "But
that means you speak Michif," he said, "Michif isn't a language, it's
just bits and pieces of other languages." However, his wife, w h o claims

One speaker, the youngest, did use some inverse forms when prompted by her
older rimtm'

Deborah Weaver


to have learned the language from him, is veryfluent,used obviative

forms on nouns, and occasionally used an obviative subject form on a
O n the other hand, those w h o are the most conservative (Group A )
are aggressively interested in the language. Three of them have been
language teachers in thefieldmethods course at the S u m m e r Institute
of Linguistics, University of North Dakota Session, for several years.
One is also a co-author of the Michif/English dictionary, which is in
preparation. Another of this group is a teacher's aide at the Ojibwa
School, an alternative school that is trying to instill a pride in the children of their heritage as Native Americans. She speaks the language
whenever she can, especially to children, and wants to help people learn
to speak it. These speakers have incorporated French nouns into the
Cree person hierarchy even though the noun phrase is in every other
way French (with the exception of the few Cree nouns). The two factors
that seem to have led to the complete loss of the proximate/obviative
distinction in the speech of the other Michif interviewed are sociolinguistic ones: a failure to use the language because of a low self-esteem
linguistically, or failure to completely learn the language as a child.
The loss of most Cree nouns and the replacement of the Cree noun
phrase with a French one has not, in and of itself, led to a loss of
the proximate/obviative distinction. Conservative speakers still retain
this distinction as seen in the responses of Group A. However, the
sociolinguistic factors surrounding the replacement of Michif by English
on the reservation may be leading to such a loss in new generations of
Michif speakers.
There is a necessity for thorough sociolinguistic research on current
language use at Turtle Mountain and its relationship to attitudes, language background, age, etc. The present study, while not a statistical
one, demonstrates that there is such a connection and lays the ground
work for further exploration of the effect of sociolinguistic factors on
the speech of the Michif at the Turtle Mountain Reservation.


The girl saw her grandmother.

Her grandmother saw her.
The boy hit his grandfather.
His grandfather hit him.
The girl saw her sister.

Obviation in Michif



Her sister saw her.

The boy hit his brother.
His brother hit him.
T h e girl saw the dog.
T h e dog saw her.
The boy hit the girl.
T h e girl hit him.


Group A


la fi u h k u m a ki:wa:pame:w (lafisu u h k u m ki:wa:pame:w)

u k u m a ki:wa:pamiku: (su u h k u m ki:wa:pamiku:)
li garso u:muluma ki:pakamahwe:w (li garso ki:pakamahwe:w su m u l u m )
u:musuma ki:pakamahuku: (su m u l u m ki:pakamahuku:)
la fi sa sor ki:wa:pame:w
sa tor ki:wapamiku:
li garso su frer ki:pakamahwe:w
su frer ki:pakamahuku:
la fi li sif ki:wa:pame:w (lafiki:wa:pame:w li licwa)
li lie ki:wa:pamiku:
li garso la fi ki:pakamahwe:w (li garso ki:pakamahwe:w la fiya)
la fi ki:pakamahuku:

Group B



la fi o h k u m a ki:wa:pame:w
ohkum(a) ki:wa:pame:w
li garso umulum(a) ki:pakamahwe:w
umulum(a) ki:pakamahwe:w
la fi sa sor ki:wa:pame:w
sa sor ki:wa:pame:w (sa sor ki:wa:pamiku:)
li garso su frer ki:pakamahwe:w
su frer ki:pakamahwe:w (su frer ki:pakamahuku:)
la fi li sif ki:wa:pame:w
li lie ki:wa:pame:w
h garso la fi ki:pakamahwe:w
la fi ki:pakamahwe:w

2 The forms in parentheses in Group A are alternative translations given by

various speakers. Note especially the presence of obviative markers on French nouns
in Sa and 6a.
* T h e forms in parentheses in Group B were given by the wife of one of the
subjects. He never used the inverse marker on his own.

Deborah Weaver


Group C



lafiki:wa:pame:w su uhkum
su uhkum ki:wa:pame:w
li garso ki:pakamahwe:w su mulum
su mulum ki:pakamahwe:w (su mulum ki:pakamahuku:)
lafiki:wa:pame:w sa sor
sa sor ki:wa:pame:w (sa sor ki:wa:pamiku:)
li garso ki:pakamahwe:w su frer
su frer ki:pakamahwe:w (su frer ki:pakamahuku:)
lafiki:wa:pame:w li lie
li lie ki:wa:pame:w
li garso ki:pakamahwe:w la fi


I am grateful to John Crawford, the director of my master's thesis at t

versity of North Dakota, for his insights and encouragement. This paper grew out
of the research that was part of that thesis. I am also indebted to Desmond Derbyshire, who read this paper in its several stages, for his comments.
Crawford, John
1076 Michif: a New Language. North Dakota English 1(4)3-10.
1078 Standardizations of Orthography in Michif. Paper presented at the
ference on Theoretical Orientations in Creole Studies. Virgin Islands.

Rhodes, Richard
1077 French Cree: a Case of Borrowing. Pp. 7-25 in Actes du Huitiime congrim
des Algonquinistes. William Cowan, ed. Ottawa: Carleton University.
Weaver, Deborah
1082 Obviation in Michif. Work Papers of the Summer
University of North Dakota Session 26:174-262.

Institute of Linguistics

The forms in parentheses in Group C were given by the older sister of one of
the subjects. The subject herself only used these forms in the presence of her sister
and reverted to the direct forms after her sister left the room.